Our project was developing an information architecture for the ibiblio branded pages. We worked with an employee of ibiblio, who was given this assignment as part of his position. In order to accomplish this task, we chose to reorganize and restructure the information on the existing webpages which serve as a wrapper and navigation for the collections hosted by ibiblio. The end result was the creation of a wireframe to present to ibiblio for review and potential implementation.

Wireframe - Wiki - User Survey - Survey Results
Sitemap - Content Analysis - Site Statistics

ibiblio is billed as "the public's library and digital archive". It began as Sunsite, hosting versions of open source software which could be freely downloaded and updated. Along the way ibiblio expanded its mission and support, and began hosting collections created by individuals who did not have the means to host the information themselves. These collections can vary from information about a single topic, to digital representation of art and primary source material, to community listserves. In recent years ibiblio has grown at an exponential rate adding hundreds of collections to their already large offering.

During this time of growth ibiblio started to redefine the purpose of the organization and how it presented itself to the public. They want to expand the idea of what they do beyond providing services, such as hosting, to include the ideas of education and research. To help communicate these changes in the organization, ibiblio felt the need to redesign the branded pages of their website.

We managed our process and meeting information on a wiki. We began having regular weekly meetings to bring everyone update on the progress we had made with the project.

We began by looking at the current site and how it was organized. The first thing we noticed was that most of the navigation found on the home page directed users to information about Linux and the open source collection. From there we began an initial content analysis and created a basic sitemap of the current pages.

We found approximately thirty pages that are maintained and branded by ibiblio. We also found from a more thorough content analysis that ibiblio was not as unorganized as we had originally thought. The organization was not apparent in our initial exploration. The goal quickly became making the organization clear and to highlight collections as well as adding in the components of research and education.

After completing the content analysis, we began organizing the information on the site by subject chunks. We went from seven chunks to three to eight. These "chunks" came to represent the global navigation in our redesign. After creating these designations and segmenting the current content into these areas, we began to think of each section as their own separate "collection". The chunks are:

• Home: Overview of ibiblio with highlights on collections and Linux as main section.
• Linux: Action oriented section with a lot of links for users to quickly navigate.
• Collections: Deep in hierarchy, the sub-navigation for this section leads to each of the categories of digital collections hosted by ibiblio.
• Streaming: A place to cleanly list radio stations and other media streamed by ibiblio.
• Events: The homepage will list events as they occur across the top. These will archive to this section for historical reference.
• Research: Current projects include Lyceum and the Linux Documentation Project. This area will grow as ibiblio strengthens this as part of their core mission.
• About: The mission, people and history of ibiblio.
• Contributors: A place for contributors to easily log in, get to the collection policy, and get help.

The next step after content analysis was to gather user information. We had difficulty with this task as ibiblio serves a worldwide audience. Our first attempt was to conduct a focus group with individuals in the local area to find out what they felt were the strengths and weaknesses of the site. After sending out emails to the webgroup through a listserv representing local users, we only received a few responses. We soon realized that this method would not work and we switched our strategy to conducting an online survey off of the site. After getting permission to add a link to the ibiblio homepage we designed our survey and built a web version.

From that method we received 66 response at the time of our presentation (the link is still available and active). From this information we were able to gather some basic information about our users.

• Most of the respondents were from the United States.
• Most of them came to ibiblio for information related to open source material but found themselves browsing the collections.
• Almost all of our respondents were comfortable with technology and were using high speed connections.

We also began to see our users divided into primarily two groups. Linux users and collection browsers. Individuals could cross over but we saw the two tasks as very different. Linux was an action oriented event want the information quickly and easily to prepare to download. Collections was a more relaxed task and encouraged exploration and browsing.

During the time we were gathering survey responses we were given site statistics as another way to look at the usage of the site. Interestingly, the number of unique users to the collections was greater than the number of users to the Linux portion of the site. Conversely, the amount of bandwidth used from the Linux portion was disproportionately larger.

Once we had more information about our users and had established our navigation we started playing around with the design. We tried several looks and had many long discussions about what would work. There was a division in the group between creating a design that was modeled after news sites such as BBC, the New York Times or NPR and a design that was modeled after digital libraries such as the California Digital Libraries, the Digital Library program of the New York Public Library or simply designed sites such as MIT.

We finally settled on a layout that was a mix of the two models. Our homepage design designates 2/3 of the space to the collections and 1/3 of the space to Linux information. The section for collections is simpler and allows for browsing and exploration while the section for Linux is action based providing several links to bring the users to the areas they most want to use. The Linux section is very flat with everything presented on the front page where as the collections section is hierarchical and offers many levels to browse.

We chose not to tackle the design aspect of the pages. After getting distracted with looks vs. information organization, we decided to not address or present color, typography or graphics. What we accomplished is an organization and architecture with concepts for page layout.

There were several major changes that we made to the overall information architecture of ibiblio. They are detailed below:

• Added global navigation, highlighting the distinct subject (or collections as we came to call them) across top which remains the same on every ibiblio branded page.
• Each collection has it's own separate home page with sub-navigation in a sidebar. The sub-navigation remains the same throughout the pages within that collection.
• The main page of each collection presents a short overview of the collection and highlights key areas for further exploration.
• The wireframe expresses proportion of page real estate and location for what is the most important idea for that particular page.
• We attempted to use the concepts of overview and preview. The homepage and main collection pages present both of these for what lies below in that section.
• An attempt was made to place the majority of the information on the homepage and each of the collection main pages above the fold.

In the end, we refined the organization but mainly didn't change the content. We attempted to present the currently displayed informatoin in a way that was more easily navigable and understood by the end user.

This was a big project to undertake for a semester and there were times when it seemed unlikely that we would be able to complete it. We began believing that what we were doing was a simple redesign which only required reworking and cleaning up to redesigning the entire mission of the site.

One lesson that was learned during the process was the vast difference between client and consultant. Throughout the process we struggled with the purpose of our project in part due to our relationship with our third team member and his connection to the organization. Although we all worked as equal team members, there was a disparity in that one member controlled all our input from our client. There were many communication breakdowns as we were all, at times, approaching the information gathering from different angles. If we had considered our third team member as a client, instead of an equal player, we may have been more thorough in gathering the information we needed about the organization and the ultimate purpose of the project.

Skills that served us well included an understanding of how to approach information organization, web design and user analysis. We all felt confident that through our past project and work experiences that we were prepared to take on this project.

As this was a work project for our third team member and not a class project, we are all invested in seeing an outcome for our creation. We've arranged to present the wireframe to ibiblio in the next week. We'd like to see them take it under consideration, audit the information to ensure that it includes all necessary information in appropriate locations. Ultimately, we'd like to see it move to a design phase with the final goal of implementation.